欧洲应对AI监管复杂性的答案是:推迟大部分条款
阅读原文· the-decoder.com欧盟已就简化版AI规则达成一致,推出《AI数字综合法案》。该法案将高风险AI系统的合规截止日期推迟至2027年底或2028年,并放宽了对中小企业的要求。法案明确禁止“脱衣”类AI应用,而针对深度伪造和AI生成文本的标注要求仍将按原计划于2026年8月生效。这一调整旨在缓解监管复杂性带来的压力。
Europe's answer to AI regulation complexity is to just delay most of it
The European Commission, the European Parliament, and the Council of the EU have agreed on simpler rules for artificial intelligence.
The deal is based on the "Digital Omnibus on AI," a legislative package that bundles several amendments to the existing AI Act. Rules for AI systems in areas like biometrics, critical infrastructure, education, and migration won't take effect until December 2027, while rules for AI in products like "lifts" or toys are pushed back to August 2028. AI systems that generate sexually explicit content without consent, including so-called "nudification" apps, are now explicitly banned.
Small and medium-sized enterprises with up to 750 employees and 150 million euros in revenue will get some relief, including reduced registration and documentation requirements along with better access to regulatory sandboxes - test environments where companies can try out AI under real-world conditions.